Detecting acoustic hot spots in closed interiors is a difficult and time demanding procedure. At the current state-of-the-art sound intensity measurements are conducted over the complete surface, to detect hot-spots where sound is transmitted into the cabin. Since this method is highly ineffective different solution approaches have been proposed to solve this problem, for example acoustic holography, inverse boundary element method and inverse finite element method. All the above mentioned methods use mathematical tools for the reconstruction of the sound field and have the difficulty that the problem is ill-posed. Therefore, all these methods need regularization techniques to find an acceptable solution. In general, the solution can be improved by using more information about the calculated system. In this paper the additional information for the regularization will be achieved by smoothing the measurement data before starting the inverse calculation. This results in a better starting point, since the smoothed measured data is closer to the correct data. The approach described in this paper models the sound field in the frequency domain using finite elements. Measurements in the interior of the domain are used as boundary conditions for the finite element analysis. The differences between the inverse finite element calculation using smoothed measurement data and not smoothed data are shown. Additionally, the influence of the smoothing on different regularization mechanisms, e.g. truncated singular value decomposition (TSVD) or Tikhonov regularization is shown.
展开▼
